Muscle has been extensively studied to identify its mechanical properties and their molecular mechanisms. The goal of our lab is to investigate how muscle properties contribute to normal motor behaviors. In posture and locomotion, the endpoint of the limb forms the primary interface with the environment. This endpoint not only needs to generate force but must also be stable to perturbations in all 3 dimensions (3D). Consider for example the case of locomotion across an uneven surface: the foot is undergoing constant 3D changes in angle and orientation and failure to deal with these changes can result in serious injury. In this proposal, we investigate the contribution of muscle properties to stability of the limb endpoint in 3D space. The key mechanical property for stability is the stiffness of the endpoint. Previous studies of endpoint stiffness have been limited by use of manipulators with only 2 degrees of freedom. Our studies utilize a 6 degree of freedom load cell coupled to a sophisticated 6 degree of freedom robotic arm to allow full exploration of endpoint stiffness in 3D space. We study the cat hind limb to allow precise measurements of individual muscle properties without intervention from reflexes. Our aims trace the sources of endpoint stiffness from single muscles, to muscle combinations, to whole limb interactions. In aim 1, we test the hypothesis that variations in muscle architecture are the primary source of differences in stiffness between muscles. In aim 2, we test the hypothesis that the force and stiffness for synergist muscles acting at single joint sum linearly. Aim 3 considers the challenge for endpoint stability generated by exertion of forces against the environment, which result in de-stabilizing reaction forces. We hypothesize that maintenance of endpoint stability during antigravity tasks actually requires co-activation of at least some flexors with the extensors that generate the primary forces. These studies will reveal how the unique design of each muscle acts to stabilize the limbs against the perturbations that actually occur during real 3D movements. These results will be important for rehabilitation engineering, providing a quantitative guide for muscle selection in tendon transfer surgeries and functional neuromuscular stimulation.